![Shrug :confusion-shrug:](./images/smilies/confusion/shrug.gif)
I sort of put myself behind the 8-ball by using mono rather than OSC, and with the speed of this object I really had no reasonable choice but to separately put together a frozen stars and frozen comet image and layer them together.
The EFW was rotated to a new filter every sub, with L getting more looks, and then a dither like every two loops through my shifting sequence.
Calibration and registration was done to a mid-track reference in PI/WBPP. Integration was then run on the star-aligned subs with strong-ish rejection, though I still had some remnant comet fuzz center frame. PI's CometAlign was then run, followed by Integration on those comet-aligned subs, with even stronger rejection to try to take out the stars and star trails, again with only moderate success.
Both stacks were then processed separately in ST, and then layer combined. The remnant comet in the one layer, and what in effect was like heavy walking noise in the other, left me stretching them both very weakly. Alas. Also, color balancing was a crap shoot. The same settings used via star sampling for the star image didn't seem to work well on the comet only image. But with all the different stacking and rejection being applied, maybe not a surprise?
It's a lot of work, all all about the rejection during integration. I'll keep trying some tricks, which hopefully will permit me to stretch both the stars only and comet only more. This comet stuff also strikes me as a good bit towards the artsy side.